What the $6 billion oil refinery upgrades mean for the Pakistani economy

Pakistan’s oil-refining problem is not simply that the country lacks capacity. It already has five refinery groups capable of processing roughly 449,400 barrels of crude oil a day, equivalent to about 20.5 million tonnes annually. The real problem is that barely half of that capacity is being used.

According to a new report by Arif Habib Limited, actual refinery throughput is only around 10 million tonnes a year, implying utilisation of less than 49%. The plants are not idle because Pakistan has no need for petroleum products. They are constrained because most were designed to produce a large quantity of furnace oil, a fuel for which domestic demand has largely disappeared.

The government’s proposed solution is a nearly $6 billion programme to upgrade the five existing refiners: Pak-Arab Refinery Company, or PARCO; Attock Refinery; National Refinery; Pakistan Refinery; and Cnergyico. The upgrades are intended to nearly double domestic petrol output, increase diesel production by almost half, sharply reduce furnace-oil production and bring the industry broadly up to Euro-V fuel standards.

That distinction is important. Pakistan does not primarily need more basic refining capacity. It needs more sophisticated facilities capable of converting a larger proportion of each barrel of crude into petrol and diesel.

The government first tried to bring about that transformation through its Brownfield Refining Policy in August 2023. Three years later, however, only Pakistan Refinery has signed a formal upgrade agreement. The remaining companies have been held back by tax disputes, financing constraints and uncertainty over whether future governments will preserve the incentives offered today.

The Petroleum Division has now proposed revised terms, including a seven-year incentive period and new investor-protection clauses. The economic prize is considerable: fewer imports of finished fuel, lower production of unwanted furnace oil, greater energy security and perhaps the largest wave of downstream petroleum investment in decades.

Yet the programme is also a reminder that announcing $6 billion of investment is easier than financing and executing it.

Pakistan’s refineries vary considerably in age, scale and technology.

PARCO and Cnergyico each account for roughly 120,000 barrels a day of capacity. National Refinery can process about 70,000 barrels a day, Attock Refinery 53,400 barrels and Pakistan Refinery 50,000 barrels.

Most of these plants are hydroskimming refineries. They can separate crude oil into its basic components and perform limited treatment, but they have much less ability than modern conversion refineries to transform heavy, lower-value products into lighter fuels such as petrol and diesel.

Some of the underlying facilities are decades old. Attock Refinery traces its origins to 1922, though its present plant and several units are much newer. Pakistan Refinery began operations in the 1960s, while National Refinery’s fuel-refining operations date to the 1970s. PARCO is the main exception, operating a comparatively more sophisticated mild-conversion refinery.

This industrial structure once made sense because furnace oil had a large domestic market. Pakistan’s power sector depended heavily on oil-fired generation, particularly after the private power policies of 1994 and 2002 encouraged investment in thermal plants.

That market has since collapsed.

Hydropower, nuclear plants, coal, liquefied natural gas and renewable generation have displaced furnace oil. Arif Habib estimates that local furnace-oil sales declined from 3.02 million tonnes in financial year 2019 to about 600,000 tonnes in 2026. Its share of electricity generation fell from 7.4% to roughly 1%.

Refineries, however, still produced furnace oil equal to around 21% of throughput in 2026. The surplus increasingly had to be exported, often at discounted international prices. Furnace-oil exports rose from virtually nothing several years ago to an estimated 1.7 million tonnes in 2026.

This creates a physical constraint. A hydroskimming refinery cannot indefinitely increase petrol and diesel production without also producing furnace oil. Once storage fills with a product domestic buyers do not want, the refinery must slow production, export at unattractive prices or shut units temporarily.

Pakistan’s refinery problem is therefore one of product mix rather than merely headline capacity.

The proposed investments are intended to convert more of each barrel into products the economy needs.

Across the sector, daily petrol production is expected to increase from 10,702 tonnes to 21,251 tonnes, a rise of about 99%. Diesel output would increase from 21,237 tonnes to 31,288 tonnes, or 47%. Furnace-oil production would fall from 15,417 tonnes to just 3,414 tonnes, a decline of roughly 78%.

The largest transformation would take place at Pakistan Refinery. Its petrol production is projected to rise from 783 tonnes a day to 4,854 tonnes, while diesel output would increase from 1,793 tonnes to 6,111 tonnes. Furnace oil would decline from 1,350 tonnes to only 167 tonnes.

Cnergyico would increase petrol output by 86% to 6,500 tonnes a day and diesel production by 29% to 11,000 tonnes, while reducing furnace oil from 7,500 tonnes to 1,000 tonnes.

PARCO would almost eliminate furnace oil, cutting output from 3,290 tonnes a day to 212 tonnes. Petrol production would rise by 49% and diesel by 44%.

National Refinery would more than double petrol output and halve furnace-oil production. Attock Refinery’s project is less dramatic in volume terms. It plans to install a continuous catalyst regeneration unit to raise petrol production by roughly one-quarter and upgrade diesel to Euro-V specifications.

These changes matter because local refineries currently meet only about 30% of Pakistan’s petrol demand and 45% of diesel demand, though they already satisfy essentially all demand for jet fuel and kerosene. Greater petrol and diesel yields would allow Pakistan to replace part of its finished-fuel imports with locally refined products.

The gross reduction in petrol and diesel imports would not translate directly into equivalent foreign-exchange savings. Refineries would need to import more crude oil, machinery, catalysts and spare parts. But the country would retain more value addition locally, reduce freight costs and become less dependent on the immediate availability of finished fuels in regional markets.

The upgrades could also provide a more reliable outlet for domestically produced crude. Refinery shutdowns or low utilisation can complicate local oil production because crude and associated natural gas are often produced together.

The most unusual part of the Brownfield Refining Policy is its financing mechanism.

Eligible refineries would receive tariff protection, known as deemed duty, in the ex-refinery prices of petrol and diesel for seven years after signing their upgrade agreements. The deemed duty on petrol would be 10%, while the incremental incentive available for diesel upgrades would effectively amount to 2.5%.

Those incentives would not simply become unrestricted refinery income. They would be deposited into escrow accounts jointly managed by the refinery and the regulator. Funds could be withdrawn only after the company achieved financial close and met specified construction milestones.

A refinery importing used equipment could withdraw escrow funds equal to as much as 24.5% of project cost. One importing new machinery could withdraw up to 27.5%. The balance would have to come from debt, equity or other financing.

In effect, the industry would use a portion of its future tariff protection to pay for its own modernisation.

The framework also contains penalties intended to prevent companies from collecting incentives without completing the investment. These include a Rs1 billion bank guarantee, third-party technical audits, penalties on delayed escrow deposits and the possible seizure of escrow accounts after prolonged default.

The revised proposal would add stability and parity clauses designed to protect investors against adverse changes in tax, regulatory or foreign-exchange policy. Such protections are particularly important for lenders financing projects that may take several years to complete and decades to repay.

Arif Habib estimates the five projects would cost a combined $5.7 billion, generally rounded to $6 billion.

Pakistan Refinery’s project is the largest at approximately $1.7 billion. PARCO’s is estimated at $1.4 billion. National Refinery and Cnergyico each require about $1 billion, while Attock Refinery’s upgrade is expected to cost around $600 million. Completion under the earlier timetable was expected between 2028 and 2030.

The escrow mechanism would finance only a minority of those amounts.

PARCO’s project is estimated to cost roughly Rs392 billion, against which it could withdraw about Rs108 billion from escrow. Pakistan Refinery would require approximately Rs476 billion and could withdraw up to Rs131 billion. National Refinery and Cnergyico would each need around Rs280 billion, with potential escrow withdrawals of roughly Rs77 billion apiece.

Attock Refinery’s Rs169 billion project would be eligible for around Rs46 billion.

Actual collections may also fall below those maximum amounts. Arif Habib estimates seven-year after-tax deemed-duty collections of around Rs175 billion for PARCO, Rs72 billion for Attock Refinery, Rs50 billion for Pakistan Refinery and approximately Rs44 billion each for National Refinery and Cnergyico.

Attock appears best placed to proceed. The company had Rs98.3 billion in cash as of March 2026 and no debt. Arif Habib assumes the project could be financed with 70% debt and 30% equity, leaving Attock well positioned to fund its equity contribution.

The research house estimates the upgrade could increase Attock Refinery’s earnings by about 18%. The elimination of penalties on lower-specification diesel, historically as high as Rs5 billion to Rs7 billion annually, would also improve profitability.

The other refiners face a harder financing task, particularly where project costs are large relative to their cash flows and balance sheets.

The original policy was notified on August 17, 2023, with a three-month deadline for signing upgrade agreements. That deadline was extended twice. Pakistan Refinery eventually signed, while the other four did not.

One obstacle was the sales-tax treatment of imported machinery. Refineries argued that taxes on billions of dollars of equipment materially damaged project economics. The FY27 Finance Act has now exempted qualifying upgrade machinery from sales tax, removing one of the principal disputes.

Financing, however, remains the larger challenge.

Pakistani banks would need to create unusually large lending consortiums. Refinery upgrades carry construction risk, exchange-rate risk, oil-price exposure and uncertainty over future pricing and taxation. The five companies could also be seeking financing at roughly the same time.

Project costs are not yet completely firm. Reliable numbers emerge only after front-end engineering design. Lenders may therefore be asked to make preliminary commitments before the final capital requirement is known, while refiners could have only six months after signing to reach financial close.

The government and the companies are consequently exploring international financing, including export-credit agencies and foreign development institutions. That search suggests domestic banking capacity alone may not be enough.

There is also the question of policy credibility. Refiners and lenders want assurances that future budgets will not impose new taxes or change pricing rules in ways that cancel out the benefits of the upgrade policy. The proposed investor protections are designed to address that concern, but they may themselves require extensive government and IMF scrutiny.

The amended policy also introduces penalties for companies that failed to sign by October 2024. In some cases, deemed duty on diesel could be reduced, increasing the cost of further delay. Yet penalties alone cannot make an unfinanceable project financeable.

The economic logic behind refinery modernisation is compelling.

Pakistan’s existing plants are poorly matched with the country’s present fuel demand. They produce too much furnace oil, too little petrol and diesel and, in some cases, fuels below the quality now required. Without upgrades, utilisation is likely to remain weak, finished-fuel imports will continue and surplus furnace oil will have to be exported at unattractive prices.

If completed, the programme could nearly double petrol output, raise diesel production by 47% and reduce furnace oil by 78%. It would improve fuel quality, strengthen supply security and make better use of industrial assets that already exist.

But the policy is not a free lunch.

Consumers ultimately fund the tariff protection. Refineries still have to raise most of the investment themselves. Much of the equipment must be imported, exposing projects to currency depreciation. Construction will take years, and returns depend on future governments honouring today’s rules.

The $6 billion figure is therefore both the attraction and the warning. It represents a potentially transformative industrial investment, but also a financing requirement far beyond what the escrow mechanism alone can provide.

Pakistan does not lack refinery capacity on paper. It lacks the capital, technology and policy certainty required to make that capacity produce what the economy actually needs.